From 9b331aeccab03f4205ab84d000cb02c4abce7af0 Mon Sep 17 00:00:00 2001
From: Mark Philipp <mphilipp622@gmail.com>
Date: Tue, 25 Aug 2026 17:58:42 -0700
Subject: [PATCH] fix(prometheus): align caller identity on latency metrics

---
 litellm/types/integrations/prometheus.py                   | 2 ++
 .../integrations/test_prometheus_caller_identity.py        | 7 ++++++-
 2 files changed, 8 insertions(+), 1 deletion(-)

diff --git a/litellm/types/integrations/prometheus.py b/litellm/types/integrations/prometheus.py
index dbd66ffaf5e..2ca299e3520 100644
--- a/litellm/types/integrations/prometheus.py
+++ b/litellm/types/integrations/prometheus.py
@@ -282,7 +282,9 @@ PROMETHEUS_DEPLOYMENT_AND_LATENCY_CALLER_IDENTITY_METRICS: Final[frozenset[str]]
         "litellm_deployment_success_responses",
         "litellm_deployment_failure_responses",
         "litellm_request_total_latency_metric",
+        "litellm_llm_api_latency_metric",
         "litellm_llm_api_time_to_first_token_metric",
+        "litellm_request_queue_time_seconds",
         "litellm_overhead_latency_metric",
         "litellm_deployment_latency_per_output_token",
     }
diff --git a/tests/test_litellm/integrations/test_prometheus_caller_identity.py b/tests/test_litellm/integrations/test_prometheus_caller_identity.py
index d7e747e8860..30c1a2e2476 100644
--- a/tests/test_litellm/integrations/test_prometheus_caller_identity.py
+++ b/tests/test_litellm/integrations/test_prometheus_caller_identity.py
@@ -277,7 +277,10 @@ def test_successful_request_emits_configured_identity_on_real_counter_and_histog
         "end_time": end_time,
         "litellm_params": {
             "custom_llm_provider": "openai",
-            "metadata": {"model_info": {"id": "deployment-id"}},
+            "metadata": {
+                "model_info": {"id": "deployment-id"},
+                "queue_time_seconds": 0.05,
+            },
         },
         "standard_logging_object": payload,
     }
@@ -304,7 +307,9 @@ def test_successful_request_emits_configured_identity_on_real_counter_and_histog
         "litellm_deployment_total_requests_total",
         "litellm_deployment_success_responses_total",
         "litellm_request_total_latency_metric_count",
+        "litellm_llm_api_latency_metric_count",
         "litellm_llm_api_time_to_first_token_metric_count",
+        "litellm_request_queue_time_seconds_count",
         "litellm_overhead_latency_metric_count",
         "litellm_deployment_latency_per_output_token_count",
     )